Servicios Personalizados
Revista
Articulo
texto en 
Español (pdf)
Articulo en XML
Referencias del artículo
Traducción automática
Enviar articulo por emailIndicadores
-
Citado por SciELO
Links relacionados
-
Similares en
SciELO
Compartir
Permalink
Revista argentina de cirugía
versión impresa ISSN 2250-639Xversión On-line ISSN 2250-639X
Rev. argent. cir. vol.116 no.2 Cap. Fed. jun. 2024 Epub 01-Ene-2024
http://dx.doi.org/10.25132/raac.v116.n2.cymvsm
Concepts and methodology
Analytical observational studies
Analytical observational studies are epidemiologic research designs in which the investigator merely observes the natural course of events without active intervention in the research subject or strict control of the environment. They are often used when it is not feasible or ethical to intervene or manipulate study variables1,2. The term analytical refers specifically to those studies that analyze and compare two groups of subjects and, if well designed, can be used to make inferences about the general population. The research design that is chosen is closely related to the research question that needs to be answered and the objectives of the study. For instance, cross-sectional studies are used to evaluate the prevalence of a condition at a specific point in time, while cohort studies are utilized to analyze the impact of various exposures, including risk and protective factors, on an event of interest.
It is important to keep in mind that observational studies do not provide strong evidence for causality, as this requires strict control of the research environment and observation of the consequences after an intervention, i.e. a clinical trial. However, they can provide valuable information about current clinical practice and serve as a starting point for designing more rigorous studies.
Observational studies must be conducted in strict accordance with ethical and methodological standards. It is essential to select an appropriate study population, define the variables of interest, standardize data collection methods, conduct long-term follow-up of patients and control for potential confounding or effect modifying factors to ensure the internal and external validity of the study. The Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines3, are a set of standardized checklists for each observational experimental design that are used to improve the quality of reporting of observational studies. These guidelines help researchers in clearly and systematically presenting the information about the study design, data collection methods, statistical analysis, and interpretation of results that is essential for the creation of useful and reliable research. In addition, in our country the use of data from human research subjects must be in accordance with the provisions of the Personal Data Protection Law 253264.
The following are the most used analytical observational research designs.
Prospective and retrospective cohort studies
In cohort studies, a group of subjects who have been exposed to a specific variable are monitored over a period to assess the occurrence of a particular event. It is important to highlight that research subjects are selected based on their exposure or not to the variable under study. When designing a study, it is essential to consider two fundamental aspects: the need for a comparison group that has not been exposed to the variable, and the requirement for the outcome to be unknown. For example, if we wanted to evaluate the effectiveness of a certain surgical technique that is commonly used in clinical practice, this design would be appropriate, but we should consider the necessity of a control group that did not undergo the technique. Furthermore, the term effectiveness should be translated into a measurable event, such as a recurrence rate or complication rate for experimental designs.
The quality of the study will be closely related to the characteristics of the exposure measurement and the event of interest. Furthermore, all potential confounding and effect modifying variables should be identified prior to data collection, and strategies for their control should be established. Another consideration is that the entire cohort should complete follow-up, and researchers should develop strategies to avoid loss to follow-up1.
Cohort studies can be classified as retrospective or prospective, depending on how follow-up is conducted. In prospective studies, the researcher collects the sample and measures exposure in the present, then follows the sample over time and observes the event of interest in the future1. In retrospective studies, a group of subjects who may or may not have been exposed in the past is identified and the event of interest is measured in the present. Follow-up occurred in the past (Table 1).
TABLE 1 Advantages and disadvantages of cohort studies
| Advantages | Disadvantages | |
|---|---|---|
| Prospective cohort | Useful to assess incidence over time It is possible to establish a temporal correlation and to estimate potential causes Measuring levels of exposure before the outcome occurs prevents the predictor measurements from being influenced by knowledge of the outcome The variables can be measured more accurately It is possible to measure incidence | Confounding or effect modifying variables may be present Are expensive and inefficient for studying rare outcomes Require large sample size |
| Retrospective cohort | Useful to assess incidence over time. It is possible to establish a temporal correlation and estimate potential causes. | Less costly and time consuming. May present biases and confounding and effect modifying variables. Require large sample size Historical data may be inaccurate or missing The researcher has limited control over the approach to sampling the population. |
When reporting the results of this design, it is advisable to use measures of association such as relative risk or odds ratio and their crude confidence intervals, adjusted for the various confounders and modifiers identified in the experimental design.
Case-control studies
In contrast to cohort studies, in this design the researcher starts with the event of interest (cases), looks for a group of people at risk for the event who are as similar as possible to the cases in all their characteristics, and measures exposure in both groups. They are useful to evaluate risk factors for rare diseases. However, this design cannot establish prevalence or incidence rates since they start from the outcome. The design of these studies is complex because it is easy to introduce selection bias in controls (Table 2).
TABLE 2 Advantages and disadvantages of case control studies
| Advantages | Disadvantages | |
|---|---|---|
| Case-control studies | Short duration First step to plan a clinical trial It is possible to measure prevalence | Cannot establish the sequence of events Not useful for rare events or exposures |
| Cross-sectional studies | Useful for studying rare outcomes Short duration and sample size | Subject for multiple biases Limited to a single event of interest The sequence of events is not clear Cannot measure prevalence or incidence |
It is recommended that results be reported using the odds ratio and its confidence interval as the measure of association. This is because the total number of events and no events (as determined by the investigator in this design) is not used in its calculation, as is the case with the relative risk. When the prevalence of the disease is less than 10%, this measure is similar to the relative risk1.
Cross-sectional studies
Cross-sectional studies are conducted at a single point in time and are useful for assessing the prevalence of a condition. They are used to evaluate the distribution patterns between variables within a population. Cross-sectional studies can also be used for examining associations, but as both the exposure and the event of interest are measured at the same time, it is difficult to establish a cause-and-effect relationship.
Other types of design
Nested case-control study
A nested case-control design has a case-control study nested within a cohort study. It is an excellent design for predictor variables that are expensive or difficult to measure and that can be assessed at the end of the study on subjects who develop the outcome during the study (the cases), and on a sample of those who do not (the controls)1,5.
Nested case-cohort studies:
This design provides an efficient alternative to cohort analyses by reducing costs while maintaining the temporal relationship between exposure and the event of interest.
In a case-cohort study, two groups are selected: a sample of individuals from the cohort (up to 100%) who have experienced the outcome of interest (cases), and 2) a (possibly overlapping) sample of individuals randomly selected from among the members of the full cohort observed at baseline (the sub-cohort).
The key to this approach is to examine the outcomes of interest in the entire sample of cases without having to collect detailed follow-up data for the entire initial cohort. This saves significant resources1,6.
Case-crossover studies:
This design is used for studying potential causes of sudden events. This design includes research subjects only with the event of interest and each case serves as his or her own control. The analysis tests whether exposure times are associated with outcome times within individuals1,7. An example of this design is the study by Lewer et al.8 which evaluated the association between opioid-related deaths in England and hospital discharge after hospital admission two years prior to death. Case-crossover designs are often statistically powerful because they allow sampling of a large proportion of cases under particular circumstances.
Referencias bibliográficas /References
1. Hulley SB, Cummings SR, Browner WS, Grady DG, Newman TB. Designing clinical research: An epidemiologic approach. 3rd ed. Philadelphia, PA: Lippincott Williams and Wilkins; 2006. [ Links ]
2. Manterola C, Quiroz G, Salazar P, García N. Metodología de los tipos y diseños de estudio más frecuentemente utilizados en investigación clínica. Rev Med Clín Condes. 2019;30(1):36-49. [ Links ]
3. Cuschieri S. The STROBE guidelines. Saudi J Anaesth. 2019;13(Suppl 1):S31-S34. [ Links ]
4. Ministerio de Justicia de la Nación. Ley de protección de datos personales. Infoleg, 30 de octubre de 2000. [Online]. https://servicios.infoleg.gob.ar/infolegInternet/ anexos/60000-64999/64790/texact.htm. [Consultado: 23 de abril de 2024]. [ Links ]
5. Ernster VL. Nested case-control studies. Prev Med. 1994;23(5):587-90. [ Links ]
6. O'Brien KM, Lawrence KG, Keil AP. The Case for Case-Cohort: An Applied Epidemiologist's Guide to Reframing Case-Cohort Studies to Improve Usability and Flexibility. Epidemiology. 2022;33(3):354-61. [ Links ]
7. Lewer D, Petersen I, Maclure M. The case-crossover design for studying sudden events. BMJ Med. 2022;1(1):e000214. [ Links ]
8. Lewer D, Eastwood B, White M, Brothers TD, McCusker M, Copeland C, et al. Fatal opioid overdoses during and shortly after hospital admissions in England: A case-crossover study. PLoS Med. 2021;18(10):e1003759. [ Links ]
This is an open-access article distributed under the terms of the Creative Commons Attribution License
